[PATCH] D80974: [DAGCombine] Adding a hook to improve the precision of fsqrt if the input is denormal

Qing Shan Zhang via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Nov 6 00:50:49 PST 2020


steven.zhang added inline comments.


================
Comment at: llvm/lib/Target/PowerPC/PPCISelLowering.cpp:12705
+  if (VT != MVT::f64)
+    return TargetLowering::getSqrtResultForDenormInput(Op, DAG);
+
----------------
qiucf wrote:
> shchenz wrote:
> > Target independent code supports vector type and we also have vector sqrt instruction on Powerpc target. Can we use them for vector types like v4f32 or v2f64?
> Yes. Supported types:
> 
> - VSX off: `f32`, `f64`
> - VSX on: `f32`, `f64`, `f128`, `v4f32`, `v2f64`
Yes, I will post another patch for the vector type as what I have mentioned in the comments of D80706.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D80974/new/

https://reviews.llvm.org/D80974



More information about the llvm-commits mailing list